Results from our Quality of Death: Ethnic and Psychosocial Influences [Coping with Cancer (CwC)] study demonstrate that at the end-of-life (EOL) black, compared with white, cancer patients receive more intensive, expensive, non-curative care (e.g., higher rates of ICU stays, ventilation, and resuscitation). We find that intensive care does not prolong life, but does impair quality of life. CwC results also reveal that black, compared to white, advanced cancer patients are less likely to consider themselves terminally ill, are more likely to prefer life-prolonging treatments, and are less likely to engage in advance care planning. In the full CwC sample, EOL discussions with physicians are associated with greater patient acceptance of terminal illness, a preference for comfort care over life-extension, fewer life-prolonging procedures, more palliative care, and receipt of care consistent with patient wishes. However, among black CwC patients, EOL discussions are not associated with greater acceptance of terminal illness, EOL care received, or EOL care consistent with patient wishes. This raises the question of how communication between oncology providers and patients may differ by whether the patient is black or white, and how these differences contribute to disparities in EOL care. For this CwC renewal, we apply the Structural Influence Model of Health Communication to determine how EOL communication processes (e.g., oncology provider's prognostic disclosure) vary depending on whether the patient is black or white. It posits that the effects of communication processes on the achievement of communication goals (e.g., acceptance of illness) will be weaker for black compared with white patients. Likewise, it posits that the effects of communication goals on EOL outcomes (i.e., intensive/palliative care received, receipt of care consistent with patient preferences, and quality of life) will be weaker in black compared with white patients. We propose to enroll 400 advanced cancer patients (200 black and 200 white patients) with a less than 6 month life-expectancy. Patients and their primary family/friend caregiver will be recruited from sites in Boston (MA), Dallas (TX), and Richmond (VA). At baseline, enrolled patients will complete a brief pre-clinic visit survey. The clinic visit will be audio-taped and patients, caregivers and the patient's primary oncology provider will be interviewed after the visit. Patients will be interviewed before and after their clinic visit one month later. That clinic visit will also be audio-taped. Medical care received in the last month of life will be documented via chart extraction. The caregiver interviewed at baseline will be interviewed one month post-loss about the health care the patient received in the final month and about the patient's mental and physical status in the last week of life. Results will indicate which communication processes and communication goals are promising targets for interventions to reduce black-white disparities in EOL care and improve quality of life for all cancer patients.